Benefits of transformer output for lab. frequency reference?

[Illusionist]

I'm designing the output stage for my 10MHz GPSDO. So far I've narrowed down the sine wave output op amps to either AD8052 or LM7171 (mostly because I have some of both).

On some designs I've seen on the 'net, the output is transformer coupled for 'isolation' (that's all that was said about it). I have some Minicircuits transformers that would suit (T1-1T-KK81+), but I don't really understand what the benefits and drawbacks are. Other than DC isolation of course, with the drawback that I then have to somehow insulate the output sockets from the (grounded) metal panel of my enclosure. Is there anything else I should know? Is it really useful? I'll be using the GPSDO as a reference to the usual suspects in my home lab.

Also, where would I put ESD protection (CM1213A I think)... on the primary or secondary side?

[edpalmer42]

In addition to the DC isolation you mentioned, transformer outputs break low-frequency ground loops (typically mains frequency) which can mess up low level or high precision measurements.

One thing to remember is that you should still have a parallel resistor/capacitor path between the shield and ground.  The capacitor gives you high frequency coupling to ground so that the shield can still act as a shield.  The resistor gives you a high value DC ground path to bleed off any charge on the shield.  I don't know if there is any concensus on what the component values should be.  One Rb standard that I know of (Efratom FRT) omitted the resistor and used 6n8 capacitors.

If you use BNC connectors, it's really easy to find cheap isolated, panel mount, female connectors.

Ed

[Illusionist]

Thank you for that. I'll stick with the transformers.

Yes, a quick search turned up some isolated BNC connectors. In fact I think I even have some, somewhere.